dynamichazard (version 0.6.6)

ddhazard_control: Auxiliary for Controlling Dynamic Hazard Models

Description

Auxiliary for additional settings with ddhazard.

Usage

ddhazard_control(kappa = NULL, alpha = 1, beta = 0, NR_eps = NULL,
  LR = 1, n_max = 10^2, eps = 0.001, est_Q_0 = F, method = "EKF",
  save_risk_set = T, save_data = T, eps_fixed_parems = 1e-04,
  fixed_parems_start = NULL,
  n_threads = getOption("ddhazard_max_threads"), denom_term = 1e-05,
  fixed_terms_method = "E_step", Q_0_term_for_fixed_E_step = NULL,
  permu = if (!is.null(method)) method == "SMA" else F,
  posterior_version = "cholesky", GMA_max_rep = 25,
  GMA_NR_eps = 1e-04, est_a_0 = TRUE, ...)

Arguments

kappa

hyper parameter \(\kappa\) in the unscented Kalman Filter.

alpha

hyper parameter \(\alpha\) in the unscented Kalman Filter.

beta

hyper parameter \(\beta\) in the unscented Kalman Filter.

NR_eps

tolerance for the Extended Kalman filter. Default is NULL which means that no extra iteration is made in the correction step.

LR

learning rate.

n_max

maximum number of iteration in the EM-algorithm.

eps

tolerance parameter for the EM-algorithm

est_Q_0

TRUE if you want the EM-algorithm to estimate Q_0. Default is FALSE.

method

set to the method to use in the E-step. Either "EKF" for the Extended Kalman Filter, "UKF" for the Unscented Kalman Filter, "SMA" for the sequential posterior mode approximation method or "GMA" for the global mode approximation method. "EKF" is the default.

save_risk_set

TRUE if you want to save the list from get_risk_obj used to estimate the model. It may be needed for later calls to e.g., residuals, plot and logLike.

save_data

TRUE if you want to keep the data argument. It may be needed for later calls to e.g., residuals, plot and logLike.

eps_fixed_parems

tolerance used in the M-step of the Fisher's scoring algorithm for the fixed effects

fixed_parems_start

starting value for fixed terms.

n_threads

maximum number of threads to use.

denom_term

term added to denominators in either the EKF or UKF.

fixed_terms_method

the method used to estimate the fixed effects. Either 'M_step' or 'E_step' for estimation in the M-step or E-step respectively.

Q_0_term_for_fixed_E_step

the diagonal value of the initial covariance matrix, Q_0, for the fixed effects if fixed effects are estimated in the E-step.

permu

TRUE if the risk sets should be permutated before computation. This is TRUE by default for posterior mode approximation method and FALSE for all other methods.

posterior_version

the implementation version of the posterior approximation method. Either "woodbury" or "cholesky".

GMA_max_rep

maximum number of iterations in the correction step if method = 'GMA'.

GMA_NR_eps

tolerance for the convergence criteria for the relative change in the norm of the coefficients in the correction step if method = 'GMA'.

est_a_0

FALSE if the starting value of the state model should be fixed.

...

additional undocumented arguments.

Value

A list with components named as the arguments.

See Also

ddhazard